Sensor Registration by Global Optimization Procedures

• US 20100076726A1
• Filed: 12/02/2009
• Published: 03/25/2010
• Est. Priority Date: 06/20/2007
• Status: Active Grant

• Current Assignee: AT&T Intellectual Property II LP (AT&T, Inc.), Raytheon Company (Rtx Corporation), University of Florida Research Foundation, Incorporated (State University System of Florida)
• Original Assignee: AT&T Intellectual Property II LP (AT&T, Inc.), Raytheon Company (Rtx Corporation), University of Florida Research Foundation, Incorporated (State University System of Florida)
• Inventors: Resende, Mauricio Guilherme de Carvalho, Hirsch, Michael Jacob, Pardalos, Panagote Miltiades
• Granted Patent: US 7,974,816 B2
• Time in Patent Office: Days
• US Class Current: 702/181
• CPC Class Codes: G06F 7/544 for evaluating functions by calculation G06F7/4824 takes precedence

Abstract

Disclosed are method and apparatus for registering multiple sensors collecting data from multiple objects. Sensor registration is decomposed into a two-step procedure. The first step corrects systematic errors. The second step assigns objects measured by one sensor to objects measured by a second sensor. Systematic errors are corrected by generating the global minimum of a systematic error function. One embodiment for generating the global minimum uses a Continuous Greedy Randomized Adaptive Search Procedure.

Claims

1. A method for registering a first sensor sensing a first set of measurements from a first plurality of objects and a second sensor sensing a second set of measurements from a second plurality of objects, comprising the steps of:

generating a likelihood function associating a data element from the first set of measurements with a data element from the second set of measurements;

generating a systematic error function based at least in part on the likelihood function;

minimizing the systematic error function; and

,

assigning an object from the first plurality of objects to an object from the second plurality of objects, based at least in part on the minimized systematic error function.

2. The method of claim 1, wherein the step of assigning uses a linear assignment method.

3. The method of claim 1, wherein the step of minimizing the systematic error function comprises:

generating a global minimum of the systematic error function.

4. The method of claim 1, wherein the step of minimizing the systematic error function uses a Continuous Greedy Randomized Adaptive Search Procedure.

5. A data processing system for registering a first sensor sensing a first set of measurements from a first plurality of objects and a second sensor sensing a second set of measurements from a second plurality of objects, comprising:

means for generating a likelihood function associating a data element from the first set of measurements with a data element from the second set of measurements;

means for generating a systematic error function based at least in part on the likelihood function;

means for minimizing the systematic error function; and

,

means for assigning an object from the first plurality of objects to an object from the second plurality of objects, based at least in part on the minimized systematic error function.

6. The data processing system of claim 5, wherein said means for assigning further comprises means for assigning using a linear assignment method.

7. The data processing system of claim 5, wherein said means for minimizing the systematic error function further comprises means for generating a global minimum of the systematic error function.

8. The data processing system of claim 5, wherein said means for minimizing the systematic error function, further comprises means for using a Continuous Greedy Randomized Adaptive Search Procedure.

9. A computer readable medium storing computer program instructions for registering a first sensor sensing a first set of measurements from a first plurality of objects and a second sensor sensing a second set of measurements from a second plurality of objects, said computer program instructions defining the steps of:

generating a likelihood function associating a data element from the first set of measurements with a data element from the second set of measurements;

generating a systematic error function based at least in part on the likelihood function;

minimizing the systematic error function; and

,

assigning an object from the first plurality of objects to an object from the second plurality of objects, based at least in part on the minimized systematic error function.

10. The computer readable medium of claim 9, wherein said computer program instructions defining the step of assigning further comprise computer program instructions defining the step of:

assigning using a linear assignment method.

11. The computer readable medium of claim 9, wherein said computer program instructions defining the step of minimizing the systematic error function further comprise computer instructions defining the step of:

generating a global minimum of the systematic error function.

12. The computer readable medium of claim 9, wherein said computer program instructions defining the step of minimizing the systematic error function further comprise computer instructions defining the step of:

using a Continuous Greedy Randomized Adaptive Search Procedure.